Hi Oliver
>I read this paper looking to see how God had managed to bias the
>results in her favour. The authors did the work in the deep south,
>tobacco growing country and tobacco is never mentioned in the
>asessment of patients admitted to the coronary care unit. However
>the unprayed over group of patients had a higher incidence of
>pulmonary disease. Could this have any significance or does prayer
>only work if you don't smoke?
>
There were 15 COPD in the prayer group and 19 in the control group - not
a significant difference and not enough to explain the difference in
outcomes. I agree it's odd the author didn't record smoking status given
the effect of smoking on CVD. However there was no significant
difference in mortality between the two groups (and I'd have liked to
see P values for each outcome cited instead of just the significant
ones). I keep going back to find methodological flaws and am unhappy
that I can't as if the effect is genuine it seriously disturbs my world
view! (of course it could be thought rays rather than God!)
I'd like to see this trial repeated elsewhere (and somehow doubt that
the effect would be repeated ;-) It's always possible that there are
hundreds of born again Christians doing similar trials, in which case a
proportion would show statistical significance even if there wasn't a
real effect - in which case this trial could be an example of
publication bias.
I'd really like to see the results of a trial of exposure to the music
of Paganini compared to the Spice Girls!
Cheers
Toby
--
Toby Lipman 7, Collingwood Terrace, Jesmond, Newcastle upon Tyne. Tel
0191-2811060 (home), 0191-2437000 (surgery)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|